Proline derivatives with functional groups on the ring carbons are useful building blocks for synthesis of pharmaceutical compounds because of the constrained conformation of proline. One such derivative, hydroxylated proline, is a starting material for the synthesis of various therapeutic compounds, including carbapenem antibiotics (see, e.g., Altamura et al., 1995, J Med Chem. 38(21):4244-56), angiotensin-converting enzyme inhibitors, protease inhibitors (see, e.g., Chen et al., 2002, J Org Chem. 67(8):2730-3; Chen et al., 2006, J Med Chem. 49(3):995-1005), nucleic acid analogs (Efimov et al., 2006, Nucleic Acids Res. 34(8):2247-2257), isoprenyltransferase inhibitors (O'Connell et al., 2000, Chem Pharm Bull. 48(5):740-742), and drug library construction (Vergnon et al., 2004, J Comb Chem. 6(1):91-8; Remuzon P., 1996, Tetrahedron 52:13803-13835). Similarly, hydroxylated derivatives of a proline homolog, L-pipecolic acid, also known as homoproline, also serve as building blocks for pharmaceutical compounds. For example, hydroxypipecolic acid is an intermediate in the synthesis of β-lactamase inhibitors (see, e.g., WO2009091856, WO2010126820 and US20110046102) and TNF-alpha converting enzyme (TACE) inhibitors (Levatic et al., 2002, Bioorg Medicinal Chem Lett. 12(10):1387-1390).
Hydroxyproline can be obtained from natural sources, such as plant materials and hydrolyzates of collagen. Hydroxyproline can also be chemically synthesized, such as from starting materials allyl bromide and diethylacetamidomalonic acid (Kyun Lee et al., 1973, Bull. Chem. Soc. Japan, 46:2924), D-glutamic acid (Eguchi et al., 1974, Bull. Chem. Soc. Japan, 47:1704-08), glyoxal and oxaloacetic acid (Ramaswamy et al., 1977, J. Org. Chem. 42(21):3440-3443), and β-alanine (Sinha et al., 2000, Proc. ECSOC-4, The Fourth International Electronic Conference on Synthetic Organic Chemistry, ISBN 3-906980-05-7).
Hydroxypipecolic acid can also be obtained from plants and other natural sources (see, e.g., Romeo et al., 1983, Phytochemistry 22(7):1615-1617; Fowden, L., 1958, Biochem J. 70(4):629-33; Clark-Lewis and Mortimer, 1959, Nature 184(Suppl 16):1234-5). Chemical synthesis of hydroxypipecolic acid is described in Callens et al., 2010, Bulletin des Sociétés Bulletin des Sociétés Chimiques Beiges 91(8):713-723; Adams et al., 1996, Chem. Commun. 3:349-350; Botman et al., 2004, Organic Letters 6(26):4941-4944; Cohen et al., 1956, Science 123(3202):842-843; Beyerman et al., 1959, Recueil des Travaux Chimiques des Pays-Bas, 78(9):648-658; Marin et al., 2004, J Org Chem. 69(1):130-41; Kumar et al., 2005, J Org Chem. 70(1):360-3; Liang et al., 2005, J Org Chem. 70(24):10182-5; Kalamkar et al., 2008, J Org Chem. 73(9):3619-22; Chiou et al., 2010, J Org Chem. 75(5):1748-51; Lemire et al., 2010, J Org Chem. 75(6):2077-80; and Angelique et al., 2000, Tetrahedron Lett. 41(36):7033-7036.
Isolation from natural sources is limited by the availability of raw materials, requires purification from a significant amount of background contaminants, and lacks certain desired diastereomers. Chemical synthetic methods can require complex steps, be difficult to scale up to industrial scale levels, and require additional purification steps due to formation of multiple hydroxylated products.
Another approach for preparing hydroxylated proline uses proline hydroxylases, which are 2-oxoglutarate-dependent dioxygenases, utilizing 2-oxoglutarate (α-ketoglutarate) and O2 as co-substrates and ferrous ion as a cofactor (see, e.g., Klein et al., 2011, Adv Synth. Catal. 353:1375-1383; U.S. Pat. No. 5,364,775; and Shibasaki et al., 1999, Appl Environ Microbiol. 65(9):4028-4031) Unlike prolyl hydroxylases that specifically recognize peptidyl proline in procollagen and related peptides, proline hydroxylases are capable of converting free proline to hydroxyproline. Several microbial enzymes that produce cis-3-, cis-4- or trans-4-hydroxyproline are known (see, e.g., U.S. Pat. No. 5,962,292; U.S. Pat. No. 5,963,254; U.S. Pat. No. 5,854,040; WO2009139365; and EP2290065) and an enzyme that produces trans-3-hydroxyproline have been identified in extracts of the fungus Glarea lozoyensis. Many of the proline hydroxylases are found in bacteria, where they are associated with the biosynthesis of peptide antibiotics. The cis-4-proline hydroxylase enzyme also shows activity in converting L-pipecolic acid (i.e., (2S)-piperidine-2-carboxylic acid) to cis-5-hydroxypipecolic acid (i.e., (2S,5S)-5-hydroxypiperidine-2-carboxylic acid; Klein et al. supra). In vitro conversions for preparing 5-hydroxypipecolic acid using these enzymes have been demonstrated, but isolated proline hydroxylases are found to denature under reaction conditions and have relatively low specific activity, rendering in vitro uses impracticable for commercial applications (Klein et al., supra). While recombinant whole cells expressing cloned proline hydroxylases are better suited for large scale industrial processes, the use of whole cells limits variations in reaction conditions, such as high substrate concentrations; restricts the types of substrates that can be used with whole cells to those that are permeable to the cells; and results in undesirable by-products that must be separated from the final product. In addition, in vivo systems may require defined growth media not optimal or cost effective because the use of rich growth media prepared from protein hydrolyzates contain free proline, which can be a competitive inhibitor when substrates other than proline are being targeted. Desirable are alternative methods for synthesizing hydroxylated forms of proline and proline analogs, as well as other chemical compounds, that can be readily scaled up and result in substantially pure stereomeric product.